INHERITANCE OF VARIATES IN THE ARMADILLO 177 



more and more bands, which give greater freedom of movement 

 without any loss of protection. Frequently, however, the body 

 armor fails to act as a single unit in inheritance, for there are cases 

 in which the bands are maternal and the other shields paternal or 

 vice versa. 



CONCLUSIONS AND SUMMARY 



What facts of importance have we succeeded in bringing to 

 light after an exacting and laborious statistical study of the data 

 presented? Have the results justified the time and labor and 

 printer's ink necessary for their publication? In reply to these 

 inquiries it may be answered (and the answer will appeal -forcibly, 

 I believe, to the reader) that the unique character of the material 

 challenges the investigator's curiosity to such an extent that he 

 is impelled to attempt to unravel the mystery. The student of 

 genetics has a right to his own opinions as to the value of these 

 results and may be able to add much by way of interpretation. 

 Whatever differences there may be regarding the theoretical bear- 

 ings of these results the data stand for themselves and are offered 

 to the expert for further analysis. 



No doubt the conclusions, for the most part, might have been 

 anticipated on the grounds of what we know of inheritance in 

 general and of the peculiar mode of reproduction of the armadillo 

 in particular. Yet it is impossible to know in advance whether 

 material will yield richly or poorly, whether we shall be rewarded 

 with striking results or only commonplace findings. 



To my mind the points that are distinctly worth the labor 

 expended in their determination are as follows : 



1. The coefficient of correlation for the polyembryonic relation 

 has been shown to be very much the same for all of the larger 

 region of the armor, in each case exceeding 0.9. We have here 

 evidence of the rather general occurrence of this unparalleled 

 approach to identity between these closest of blood relatives. This 

 finding strongly supports the taxonomic assumption that degrees 

 of resemblance are in direct ratio to degrees of blood relationship. 



2. Within single sets of quadruplets there is a sort of segrega- 

 tion or distribution of the parental influences, so that some indi- 



THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 15, NO. 2 



